OK, here's my design.
The circuit uses two ICs, both from the CMOS "CD4000" series. These are old technology but still available and still used nowadays. U1 is a CD4093B aka HCF4093B, and is shown as four separate "D"-shaped items on the schematic, named U1A~U1D, called "NAND gates". U2 is a CD4024B aka HCF4024B. The prefix differs between manufacturers, that's all.
Both U1 and U2 have "decoupling" capacitors (CD1 and CD2) connected across their VDD and VSS pins (pins 14 and 7). These must be connected as directly as possible to those pins of the IC (or the socket, if you use sockets, which you should). It's best to put them on the underside of the board, soldered directly between those corner pins, with plastic sleeving (stripped off some hookup wire) slid onto their wires so they don't short onto any of the other copper. These decoupling capacitors are needed to ensure reliable operation of the ICs. Google decoupling capacitors for more information.
The circuit consists of, from left to right, an oscillator, to generate the on-off-on-off signal that provides the flashing, then a circuit that generates the 8+8 control signals, then a three-position toggle switch that selects between ON, OFF, and flashing, then two MOSFET drive stages and two power MOSFETs that switch the current to the LED arrays.
Connections are all made through the six-pin connector at the right.
All components are available from Maplin and part numbers are listed at the bottom. You will also need a piece of prototyping board - Maplin have several options:
FL17T 21x4 cm
http://www.maplin.co.uk/p/veroboard-copper-dil-stripboard-381x2146mm-fl17t
JP47B 10x7.4 cm
http://www.maplin.co.uk/p/2939-srbp-matrix-board-strip-jp47b
JP48C 15x7.4 cm
http://www.maplin.co.uk/p/2958-srbp-matrix-board-strip-jp48c
JP50E 16x10 cm
http://www.maplin.co.uk/p/3962-srbp-matrix-board-strip-jp50e
JP51F 30x10 cm
http://www.maplin.co.uk/p/39117-srbp-matrix-board-strip-jp51f
Any of those should be large enough for the whole circuit, since there are no heatsinks needed, and no really large components. The first one has the tracks cut in two, which is convenient for mounting the two ICs. If you plan on doing more electronic construction, you could get a larger board. It's easy to snap it to size.
You may also want to get a "spot face cutter" for cutting stripboard tracks (
http://www.maplin.co.uk/p/spot-face-cutter-fl25c) - or you can just use a 3.5 mm drill bit.
You should use IC sockets - two of these:
http://www.maplin.co.uk/p/dual-in-line-socket-14-pin-bl18u. The ICs can be damaged by static electricity so it's best to build the circuit around the sockets, and install the ICs at the last minute.
The proper part number for RF is not clear from Maplin's description, and their photos aren't helpful either. The numbers I've listed are for a "vertical mount" one and a "horizontal mount" one. Get one of each, and use the one that can be adjusted directly from above.
Q3 and Q4, the power MOSFETs, can be any of the types I've listed on the schematic. They are listed with the best option first. All of those are available from Digi-Key, but the last option, IRF540, is the only power MOSFET that Maplin carry! It is suitable but it may get slightly warm during operation. These MOSFETs come in a TO-220 package that can be mounted onto a heatsink, but no heatsink is needed in this application. You can stand them up vertically on the board, or lie them flat against the board and screw them to it, if there's enough room.
MOSFETs, like CMOS ICs, can be damaged by static electricity during handling and soldering. If possible, push the leads through a piece of conductive foam, or hook a paper clip around them, while you're installing them into the board. (Install them last; just before the ICs.)
The circuit is designed to run on an automotive supply of nominally 12V, minimum 10V, maximum 15V. The LED arrays are driven with the full input voltage (minus a small voltage drop in the MOSFETs - less than 0.1 volts). The output MOSFETs can supply at least 5A into each LED array.
The right hand section of the schematic contains several thick lines. These indicate the high current paths in the circuit. You should ensure that these paths are all made with wire; do not rely on the copper of the stripboard for any of those parts of the circuit. The rest of the circuit operates at low current and doesn't need those precautions.
Here's a circuit description in case you're interested.
U1 contains four identical functional blocks. Each one is called a "NAND gate with Schmitt trigger inputs" and they're quite versatile. Each one has two inputs (on the left) and one output, and it drives its output high unless both of its inputs are high.
"High" and "low" refer to voltage, which is measured relative to the 0V rail that runs along the bottom of the diagram. A voltage is considered high if it's between about 8V and 12V, and low if it's between 0V and 4V. (In between 4V and 8V it is undefined.)
U1A is used as an oscillator, and its "Schmitt trigger input" feature is essential for this function. I won't go into detail here, but search for 4093 Schmitt trigger oscillator, either here on Electronics Point or using Google, if you want to learn more.
U1A produces a voltage that alternates (oscillates) between high and low with about three high-low cycles every second. The actual rate is determined by CF and RF, and can be adjusted by RF, which is a preset potentiometer or "trimpot" - an adjustable resistance that you turn using a screwdriver to set the flash rate you want.
This signal feeds into U2, which is a "ripple carry binary counter". It consists of a chain of cascaded "toggle flip-flops"; each flip-flop produces an output signal that toggles (changes from low to high, or from high to low) each time its input changes from high to low. So the output will go through one complete high/low cycle when the input goes through TWO high/low cycles. So the output changes at half the rate of the input.
Each of the Q outputs on U2 changes at half the rate of the one before it. So, Q1 on pin 12 goes high and low eight times before Q5 on pin 5 changes state.
Q5 on pin 5 determines which LED array the flashes are directed to. This is done by the remaining three NAND gates in U1. First, U1D inverts the Q5 signal. In other words, its ouput is the opposite of the Q5 signal. Then the flash signal from Q1 (pin 12 of U2) feeds into both U1B and U1C, while the other inputs of those gates are driven from Q5 and inverted Q5, respectively.
A NAND gate drives its output high, unless both of its inputs are high, in which case it drives its output low. The resulting signals on pins 4 and 10 of U1 consist of eight pulses, then an equal gap during which the signal is high. This would be easier to describe with a diagram; let me know if you'd like me to draw one.
These are the flashing control signals for the two set of LEDs. A low signal corresponds to the LEDs being ON. They are passed through switch SW1A, which has three positions, and on to R1 and R2. In the top position (as drawn in the schematic), the signals on R1 and R2 follow the flashing pattern I just described.
In the middle position, R1 and R2 aren't connected to anything - this is how three-position toggle switches normally work; in the middle position, nothing is connected to anything - that's why these switches are called "ON-OFF-ON" switches. In the bottom position, R1 and R2 are connected to 0V, i.e. low, which corresponds to LEDs ON.
R1 and R2 feed these states into Q1 and Q2 which drive Q3 and Q4, the output MOSFETs. Both of these driver and output stages are the same so I'll only describe one. When the input at R1 is low, current flows through R1 and into Q1's base, turning Q1 ON. Q1 pulls its collector high and this provides positive voltage through R7 onto Q3's gate, which causes Q3 to conduct. Q3 completes the circuit and applies voltage across the LEDA array.
Current flows from the +12V supply on CN1 pin 1, out CN1 pin 2, through the LED array, back into CN1 pin 3, into Q3's drain, through Q3, out Q3's source, onto the 0V rail, out CN1 pin 6 and back to the return lead of the supply.
When R1 is high (or not connected to anything, if SW1 is in the "OFF" position (the middle position), no current flows into Q1's base, Q1's collector is pulled low by R5, there is no gate bias on Q3, and Q3 does not conduct any current into the LED array.
R9 and DP provide protection against automotive load dump and other disturbances on the power supply. Q3 and Q4 are rated for 100V so they can withstand the load dump voltage when they're OFF, but it could damage the LED panels. Google load dump for more information.
